Endocrine Abstracts

Evaluation of male fertility, based on semen analysis, is complicated by high variability of semen parameters and the lack of evidence-based reference values. The standards are currently set by the WHO laboratory handbook (1999). We investigated whether taking into account abstinence time and testicular volume would reduce variation in sperm concentration / count and thereby consolidate the underlying principles of spermatogenesis.

The database of our Institute provided data from 1976 to 2005 on 30,965 semen samples from 11,062 infertile men and their abstinence times, testicular volumes and hormone levels. Their impact on semen parameters was analysed by correlation with each individual factor alone and multiple regression analysis.

A highly significant linear correlation between duration of abstinence and sperm concentration/total sperm count was found. Total sperm counts increased between 2.4 (1 d abstinence) and 6.9×10⁶ (8 d of abstinence) per ml of bi-testicular volume. Sperm concentration varied with semen volume, which exhibited a linear increase with abstinence but reached a plateau after 8 d. The percentage motility and normal morphology changed only marginally between 1 and 8 d of abstinence. Total sperm counts were inversely related to serum FSH but not to testosterone concentrations.

Conclusions: 1. Correlations between total sperm count and duration of abstinence, testicular volume and FSH are stronger than with sperm concentration, so total sperm count is the more reliable parameter for indicating testicular function. 2. Clinical evaluation of semen samples should take abstinence and testicular volume into account and the diagnostics can be further improved by assessing FSH levels. 3. Normal values for semen analyses need to be established for different abstinence times and testicular volumes. 4. Recommendations for short abstinence periods before timed intercourse or assisted reproduction are not reasonable, as total sperm counts increase far more than motility and morphology vary.